Você está na página 1de 5

Fast Fission Assisted Ignition of Thermonuclear Microexplosions

Friedwardt Winterberg
University of Nevada, Reno, USA
Reprint requests to Prof. F. W.; E-mail: winterbe@physics.unr.edu

Z. Naturforsch. 61a, 559 – 563 (2006); received June 27, 2006

It is shown that the requirements for fast ignition of thermonuclear microexplosions can be sub-
stantially relaxed if the deuterium-tritium (DT) hot spot is placed inside a shell of U-238 (Th-232). An
intense laser – or particle beam-projected into the shell leads to a large temperature gradient between
the hot DT and the cold U-238 (Th-232), driving thermomagnetic currents by the Nernst effect, with
magnetic fields large enough to entrap within the hot spot the α -particles of the DT fusion reaction.
The fast fission reactions in the U-238 (Th-232) shell implode about 1/2 of the shell onto the DT,
increasing its density and reaction rate. With the magnetic field generated by the Nernst effect, there
is no need to connect the target to a large current carrying transmission line, as it is required for mag-
netized target fusion, solving the so-called “stand off” problem for thermonuclear microexplosions.
– PACS number: 28.52.-s.

Key words: Fast Ignition; Inertial Confinement Fusion.

1. Introduction this concept to work a petawatt laser with a total en-

ergy output of more than 100 kJ is needed. Taking into
One of the more promising inertial confinement consideration the low efficiency of petawatt lasers, this
thermonuclear fusion (ICF) concepts, going by the appears to be a very expensive proposition, and it again
name of the fast ignitor [1], proposes isentropic com- raises the question of whether there might be a much
pression to high densities of a solid deuterium-tritium less expensive way to achieve the goal of controlled
(DT) target, followed by hot spot ignition with an ul- nuclear fusion by inertial confinement. I claim it ex-
trashort petawatt laser. Estimates indicated a drastic re- ists in the coupling of the DT fusion fuel with a com-
duction of the total energy needed for compression and paratively small amount of natural uranium (U-238) or
ignition to perhaps as little as ∼ 100 kJ. The compres- thorium (Th-232).
sion could be done as in other ICF concepts by laser or Large thermonuclear explosive devices are ignited
particle beams [2 – 5] or even by soft X-rays from elec- by the explosive energy from a critical mass of U-235
tric pulse powered imploded arrays of thin wires [6]. (Pu-239). Not making extravagant such a device, the
For the fast ignitor, two petawatt lasers were originally thermonuclear fusion energy released must be at least
proposed. The first one was to bore a hole into the cen- as large as the energy released by fission. But because
ter of the compressed DT, with the second one deliver- such a device depends on a critical mass of several
ing the energy needed for ignition to the center of the kilograms at least, this kind of “fast ignition” cannot
DT through the hole opened by the first laser. Later, an be scaled down for the ignition of thermonuclear mi-
important modification of the original fast ignitor con- croexplosions. However, for small thermonuclear as-
cept was to replace the first laser needed to bore the semblies a different way to release the energy from a
hole by a hollow gold cone stuck in the DT with the fission reaction, in conjunction with a DT (DD) fusion
tip of the cone in the center of the DT [7]. Theoreti- reaction, can be used profitably, greatly relaxing the re-
cal and experimental studies showed that the inertia of quirements for fast ignition. For this process even natu-
the gold cone appears to be sufficient to keep the cone ral uranium (U-238) and thorium (Th-232) can be used,
open long enough for the laser-ignition pulse to reach and, as for large thermonuclear explosive devices, the
the center of the DT [8]. relative amount of the fission product fall-out can be
Apart from the complexity of the fast ignitor ap- kept quite small. The importance of this possibility
proach towards ICF, aggravated by the smallness of the is, that it greatly increases the chance for controlled
DT target assembly, the consensus has emerged that for nuclear fusion to succeed, justifying the price which

0932–0784 / 06 / 1000–0559 $ 06.00 

c 2006 Verlag der Zeitschrift für Naturforschung, Tübingen · http://znaturforsch.com
560 F. Winterberg · Ignition of Thermonuclear Microexplosions

Fig. 1. Bombardment, by a laser or particle beam LB, of a spherical solid DT target surrounded by a natural uranium (thorium)
shell, with propagating burn into a DT cylinder inside a liner L.

has to be paid in the release of a comparatively small into a DT cylinder. There too, a metallic cylindrical
amount of fission products. shell surrounding the DT cylinder can set up a large
magnetic field by the Nernst effect, making possi-
2. Fast Fission Assisted Hot Spot Fast Ignition ble a magnetic field assisted thermonuclear detonation
Assembly wave propagating down the cylinder. In addition, soft
X-rays, released from both the hot spot and the rear
As explained in Fig. 1, a sphere containing DT, sur- of the detonation wave, can be utilized to compress the
rounded by a shell of metallic U-238 (Th-232), is po- un-burnt DT ahead of the wave. Because the amount of
sitioned at one end of a cylindrical thermonuclear mi- U-238 (Th-232) needed to create the hot spot is small
croexplosion assembly. The DT inside the shell can be in comparison to the total amount of DT, the amount
heated through an opening to thermonuclear temper- of the undesirable fission products is relatively small.
atures by a laser or particle beam. A large tempera-
ture gradient between the hot DT plasma and the cold 3. The Thermomagnetic Nernst Effect
U-238 (Th-232) shell surrounding the DT plasma gen-
erates near the DT-U-238 interface currents by the ther- A temperature gradient in a magnetized fully ion-
momagnetic Nernst effect with magnetic fields large ized plasma leads to thermomagnetic currents (Nernst
enough to entrap the α -particles from the DT fusion re- effect), with the current density for a hydrogen plasma
action in the DT plasma. In passing through the U-238 given by [9]
(Th-232) shell, some of the 14 MeV DT fusion reac- 3nkc
tion neutrons perform fast fission reactions in the shell, jN = H × T, (1)
2H 2
heating it to high temperatures. If the thickness of the
valid for ωτ  1, where ω is the electron cyclotron
shell is properly chosen, the shell will implode onto the
frequency and τ the electron-ion collision time. In (1)
DT increasing the density and reaction rate of the DT
n is the number density of electrons and protons, H the
plasma inside the shell, further accelerating the implo-
magnetic field strength, T the temperature gradient;
sion of the shell and with it the DT thermonuclear re-
k and c are the Boltzmann constant and the velocity
action rate, in an “autocatalytic” fusion-fission-fusion
of light. The current density (1) has a magnetic body
force density acting on the plasma equal to
To set up the magnetic field by the currents of the
thermomagnetic Nernst effect, a seed field is needed, 1 3 nk
f = jN × H = (H × T ) × H. (2)
which is amplified by the Nernst effect. The seed field c 2 H2
can be provided by placing the entire assembly in the If T ⊥ H, this becomes
center of a large magnetic solenoid, with the axis of the 3
microexplosion assembly in the same direction as the f = nkT. (3)
direction of the magnetic field.
Inserting (3) into the magnetohydrostatic equation
From the fission assisted fusion hot spot thus pro-
duced, a thermonuclear detonation wave is launched p= f (4)
F. Winterberg · Ignition of Thermonuclear Microexplosions 561

and setting p = 2nkT , valid for a hydrogen plasma, one with a = 2.7 · 10 5 Gcm is the Larmor radius of the α -
obtains from (4) particles. For H = 2 · 10 8 G, one has rL ∼ 3 · 10−3 cm,
smaller than r0 by more than two orders of magnitude.
n T + 4T n = 0 (5) Finally, the electron Larmor frequency at H = 2 · 10 8 G
is ω ∼= 3.5 · 1015 s−1 , and τ ≈ 6.7 · 10 −14 s, hence
or ωτ ≈ 230  1.
The return current is going in the opposite direc-
T n4 = const. (6)
tion along the cold surface of the U-238 shell, where
The Nernst current density is here directed azimutally ωτ  1, with no thermomagnetic EMF in the opposite
and is direction. Accordingly, the magnetic lines of force are
closed, forming a torus wound around the z-axis.
3nkc dT
jN = . (7)
2H dr 4. Autocatalytic Fusion-Fission-Fusion Reaction
Inserting jN into Maxwell’s equation
The concept of the “autocatalytic” fusion-fission-
4π fusion reaction previously proposed [10] is here pre-
jN = ×H, (8) sented in a simplified way [11, 12]. The DT thermonu-
clear reaction rate inside the U-238 shell is determined
one obtains with H directed along the z-axis by the equation
dT dH
6π nk = −H . (9) ∂n n2
dr dr =
σ v , (12)
∂t 4
With T n4 = T0 n0 4 , where T0 and n0 are the temperature
σ v is the DT nuclear fusion reaction cross sec-
and number density at r = 0, (9) becomes
tion σ multiplied with the particle velocity v and av-
eraged over a Maxwellian. With the help of (12) one
−Hd H = 6π n0 kT0 T −1/4 d T.
obtains for the neutron flux at the surface of the DT
Integrating (10) from H 0 , T = 0 at the cold wall made sphere of radius r 0
from U-238 to H = 0, T = T0 at r = 0, one obtains r0 2
1 n
H02 Φ=
σ v 4π r2 dr. (13)
= 2n0 kT0 . (11) 4π r02 4
8π 0

The return current flows along the cold wall with the Assuming that n2
σ v does not depend on r, one finds
magnetic field of the Nernst current shielding the inte- that
rior of the hot DT plasma from the magnetic field of
the Nernst current. n2
σ v r0 . (14)
As an example, let us take a sphere of solid DT with 12
n0 = 5 · 1022 cm−3 and with a radius r0 = 0.1 cm. To
The fission reaction path length of the neutrons in
be heated to T0 = 108 K, the ignition temperature of
U-238 is Σ−1f , where Σf = nf σf is the macroscopic
the DT reaction requires the input energy
fission cross section where σf is the fission reaction
Ein = 2n0 kT0 (4π /3)r03 = 6 · 1012 erg = 600 kJ. cross section and n f the atomic number density. For
metallic U-238 one has n f ≈ 4 · 1022 cm−3 and σf ≈
According to (11), the magnetic field at the 2 · 10−24 cm2 , hence Σ−1
f ≈ 8 cm.
√ plasma boundary at r = r 0 would be H0 = Assuming that the thickness of the U-238 shell is
16π n0kT0 ∼= 2 · 108 G, with a Nernst current IN = δ  Σ−1f , then only the fraction δ Σ f  1 of the neu-
5r0 H0 = 10 A. To deflect and entrap the α -particles
8 trons passing through the shell will make a fission re-
from the DT fusion reaction inside the sphere of radius action. Because this fraction is small, the neutron flux
r = r0 = 0.1 cm requires that r L  r0 , where rL = a/H over the shell remains approximately constant.
562 F. Winterberg · Ignition of Thermonuclear Microexplosions

The energy released per unit volume in the time τ in finds that r0 δ |min ≈ 2 · 10−2 cm2 ; or for r0 = 0.1 cm,
the U-238 shell is that δ ≈ 0.2 cm.
We can now also compute the implosion velocity of
ε = Σf Φ (εf + ε0 )τ , (15) the U-238 shell. For kT0 = 15 keV = 2.4 · 10 −8 erg, the
pressure in the DT plasma with n = 5·10 22 cm−3 is p =
where εf is the fission energy released and ε 0 the ki- 2nkT0 = 2.4 · 1015 dyn/cm2. For the autocatalytic im-
netic energy of the neutrons released from the DT reac- plosion to take place, the pressure (i. e. energy density)
tion. τ is the time of inertial confinement of the U-238 in the U-238 shell must be at least that large. Therefore,
shell. By order of magnitude it is equal to with ρ = 18 g/cm3 for U-238, the implosion velocity
τ ∼ δ /a, (16) is vimp = p/ρ ≈ 107 cm/s. And with δ = 0.2 cm, the
inertial confinement time is τ ≈ 2 · 10 −8 s. Therefore,

where a ≈ p/ρ is the velocity of sound in the hot nτ ≈ 1015 cm−3 s, well about nτ ≈ 10 14 cm−3 s (Law-
U-238 shell of density ρ . son criterion) to assure an exponential growth of the
The justification of (15) can perhaps be better seen fusion-fission autocatalytic reaction.
by writing it with the help of (16) as
  5. Propagating Thermonuclear Burn from the
δ 1 Fission Assisted Fusion Hot Spot
ε = Φ (εf + ε0 ) , (15a)
L a
Once a fast fission assisted hot spot is created, a
where L = 1/Σf is the fission reaction path length in propagating thermonuclear burn along a slender DT
the shell, with the fraction (δ /L) of neutrons passing cylinder is possible with the aid of a large axial current
through the shell making a fission.
 Without the division along the cylinder [13] established here by the Nernst
by the velocity of sound, a = p/ρ , (15a) would be effect. As shown in Fig. 1, the efficiency of the burn
the energy flux density in the shell. can be further improved by precompressing the DT
Since by order of magnitude p ≈ ε one has fuel cylinder with the soft X-rays coming from the hot
 spot and the rear of the burn wave. In this way a large
τ = δ ρ /ε (17) thermonuclear yield seems possible, large in compari-
son to the small amount of energy released by the fast
and one obtains from (15) fission.
ε = [Σf Φ (εf + ε0 )δ ] /3 ρ 1/3 .
6. Conclusion
By inserting Φ from (14), this becomes
In fissionless fast ignition concepts the energy re-
1  2/3 1/3 quired for fast ignition is of the order ∼ 10 5 J, to
ε = 2/3 Σf (εf + ε0 )
σ v n2 r0 δ ρ . (19) be delivered in ∼ 10 −10 s by an expensive inefficient
∼ 1015 W laser. In addition, a ∼ 10 6 J – 1014 W laser
The shell begins to implode onto the DT if ε > 2nkT . pulse is there required for compressing the target to
With the help of (19), this condition can be written as thousand-fold solid density. By comparison, in the fis-
sion assisted fast ignition concept proposed here, the
(kT )3 1152
n(r0 δ )2 > . (20) total energy is estimated to be about 6 · 10 5 J, to be de-

σ v [Σf (εf + ε0 )]2 ρ
livered in 2·10 −8 s by a much less expensive 3·10 13 W
The right hand side of this equation is smallest for the laser, with the target compression done “autocatalyt-
minimum of (kT ) 3 /
σ v 2 , which is at kT ≈ 15 keV, ically” by the coupling of the fusion with the fission
where (kT )3 /
σ v 2 ≈ 2 · 108 erg3 s2 /cm6 . We therefore process, and with no need to compress the target to
have thousand-fold solid density. With the relatively small
fission reaction “fall-out”, the proposed concept may
n(r0 δ )2 |min ≈ 2 · 1011[Σf (εf + ε0 )]−2 ρ −1 [cm]. (21) have distinct advantages over pure fusion inertial con-
finement microexplosion concepts.
For U-238 with Σ f ≈ 0.8 · 10−1 cm−1 , εf + ε0 = 3 · Finally, I may mention that the idea of a natural ura-
10−4 erg, ρ = 18 g/cm 3 , and for n = 5 · 10 22 cm−3 , one nium shell surrounding a solid DT sphere with the DT
F. Winterberg · Ignition of Thermonuclear Microexplosions 563

ignited by a relativistic electron beam was proposed in clear microexplosion rocket propulsion system, but the
1971 [14] as a means for an efficient Orion-type nu- concept was not further explored.

[1] M. Tabak, J. Hammer, M. E. Glinsky, W. L. Kruer, S. C. R. G. Evans, H. Fujita, Y. Kitagawa, K. Krushelnick,

Wilkes, J. Woodworth, E. M. Campbell, and M. D. T. Miyakoshi, N. Miyanaga, T. Norimatsu, S. J. Rose,
Perry, Phys. Plasmas 1, 1626 (1994). T. Shozaki, K. Shigemori, A. Sunahara, M. Tampo,
[2] J. Nuckolls, in: Laser Interaction and Related Plasma K. A. Tanaka, Y. Toyama, T. Yamanaka, and M. Zepf,
Phenomena (Eds. H. J. Schwarz and H. Hora), Plenum Nature 412, 798 (2001).
Press, New York 1974, Vol. 3B, p. 399ff. [8] R. Kodama and the Fast-Ignitor Consortium, Nature
[3] K. A. Brueckner, in: Laser Interaction and Related 418, 933 (2002).
Plasma Phenomena (Eds. H. J. Schwarz and H. Hora), [9] L. Spitzer, Physics of Fully Ionized Plasmas, 2nd ed.,
Plenum Press, New York 1974, Vol. 3B, p. 427ff. Interscience Publishers, John Wiley & Sons, New York
[4] R. Martin, IEEE Trans. Nucl. Sci. 22, 1763 (1975). 1962, p. 145.
[5] A. W. Maschke, IEEE Trans. Nucl. Sci. 22, 1825 [10] F. Winterberg, Atomkernenergie – Kerntechnik 44, 145
(1975). (1984).
[6] T. W. L. Sanford, T. J. Nash, R. C. Mock, R. B. Spiel- [11] F. Winterberg, Z. Naturforsch. 58a, 612 (2003).
man, K. W. Struve, J. H. Hammer, J. S. De Groot, K. G. [12] F. Winterberg, Phys. Lett. A 336, 188 (2005).
Whitney, and J. P. Apruzese, Phys. Plasmas 4, 2188 [13] F. Winterberg, Atomkernenergie – Kerntechnik 39, 265
(1997). (1981).
[7] R. Kodama, P. A. Norreys, K. Mima, A. E. Dangor, [14] F. Winterberg, Raumfahrtforschung 15, 208 (1971).